I'm posting a 5D2 shot from the other night, in circumstances that I believe would challenge any digital sensor. Black night to sunlit rock in the same shot, with lots of subtle colour tones. Subtlety like this in such extreme situations was impossible with film IMHO. I'm so impressed with the 5D2 sensor, really.

5D2 is an amazing camera

IIRC Most DSLR's share roughly the same DR and are no match for film. But I could be wrong

"Most DSLR's share roughly the same DR and are no match for film"

Interesting, I've tried shots like this for years with film, when I got the colours in the clouds the moon would be completely obliterated by halation. Makes me think that there is more to DR than is commonly described. I don't know how to put it, but the 5D2 is allowing me to do stuff that I couldn't do with film.

Raw thesedays out of most SLRs is 11,12 or even 13 stops DR, prehaps not all usable but still. Film is surpassed, completly.

RobDickinson wrote:
Raw thesedays out of most SLRs is 11,12 or even 13 stops DR, prehaps not all usable but still. Film is surpassed, completly.

Yeah, many have this bizarre notion that print film is vastly superior to digital still. Often I hear that digital has "6-7 stops" range and color print film "14-15". Neither is even remotely correct in my experience.

Now, well processed black and white film (e.g. Tmax 100) I think still has the edge by a fair margin.

What is film

^^^

saaketham wrote:
What is film

It's something that we use to wrap sandwiches.

David Baldwin wrote:
"Most DSLR's share roughly the same DR and are no match for film"

Interesting, I've tried shots like this for years with film, when I got the colours in the clouds the moon would be completely obliterated by halation. Makes me think that there is more to DR than is commonly described. I don't know how to put it, but the 5D2 is allowing me to do stuff that I couldn't do with film.

I did say I could be wrong...

That's because it has been a while since I used color negative film. Still, IIRC when I shot a high contrast scene with a simple color film, I didn't get blown out skies or blocked shadows that easily (compared to digital). With digital I always have to choose between sacrifice detail in the shadows or blow out highlights. A tough decision. Especially since the 5D2 shadows aren't very pretty when you open them up. Effectively reducing it's DR or at least the possibilities of enlarging DR artificially in PP. Although de-banding software like Dfine2 does a great job at removing banding (at the expense of some detail).

Anyway, I am not impressed with the 5D2's DR myself. I think it is usable, but nothing special. OTOH I did use a Nikon D90 that would take a +50 fill light adjustment in LR without pain. The 5D2 files start to fall apart at around +20 fill light adjustment. I was really impressed with the D90... Sadly I sold it to fund a 5D2

compared to slides, the DR of sensors is quite good, negative film delivered better DR however. The advantage of DSLR is that you have a Pola back to check your image and that it is easy to merge shots for more DR. Hi ISO and hi DR is also something which is amazingly good with the 5DII

Daan,

Interesting point. I've been following the 5D2 shadow debate with interest. The question for me is whether DR is the natural range of the sensor without heavy PP (which my moon picture relates to), or is DR how much PP work the image will stand before it falls to pieces.

Don't know the answer to this, but I can't help but feel that some of the complaints about the 5D2 shadow noise are really about poor exposure, not true DR. Perhaps I am being unfair?

To be clear, in the original RAW file of my moon image the outer clouds are black, and show no banding. I visualized those clouds as being black when I chose the exposure. If I then tried to "open up" the clouds by a few stops to make them gray in post, and had banding, wouldn't I be being unreasonable? If the camera file records black, shouldn't I expect that tone to print black, and then be happy?

With film photography we all tried hard to visualize our tones when we chose our exposure, surely that visualization process should apply to digital too. Exposure to the right is a great rule of thumb, but in terms of good RAW reproduction I don't believe it is the last word. There must be some sort of zone like system applicable to RAW, but I can only start guessing as to what that process might be.

David Baldwin wrote:
Don't know the answer to this, but I can't help but feel that some of the complaints about the 5D2 shadow noise are really about poor exposure, not true DR. Perhaps I am being unfair?



Dave,

Aside from the fact that problems at low-ISO without pushing the files are present (see the night shot noise on page 3 of the 'show us your shots the 5DII wrecked thread) The point many on the "Pro-5DII" side don't talk about is the fact that a regular 5D, the camera the 5DII was supposed to improve upon, is so much more flexible to be pushed in post without pattern noise. If they could do it before, why not now?

David Baldwin wrote:
Daan,

Interesting point. I've been following the 5D2 shadow debate with interest. The question for me is whether DR is the natural range of the sensor without heavy PP (which my moon picture relates to), or is DR how much PP work the image will stand before it falls to pieces.

Don't know the answer to this, but I can't help but feel that some of the complaints about the 5D2 shadow noise are really about poor exposure, not true DR. Perhaps I am being unfair?

Not unfair, but misunderstood. You can exposure a high contrast scene (like your moon shot) perfect for the highlights and end up with blocked up shadows (because of the limited DR). So, if you want to have more detail visible in your shadows, you will have to open them up (with the fill light tool for example). In effect, how far you can open the shadows up with such a tool before IQ degrades, determines the (artificial) DR. You could expose for the shadows too... but that would mean you would blow the highlights. And blown highlights are much more difficult to bring back (and look ugly/digital when printed) than opening up shadows.

To be clear, in the original RAW file of my moon image the outer clouds are black, and show no banding. I visualized those clouds as being black when I chose the exposure. If I then tried to "open up" the clouds by a few stops to make them gray in post, and had banding, wouldn't I be being unreasonable? If the camera file records black, shouldn't I expect that tone to print black, and then be happy?

If your cam would have had more DR, the black clouds would have contained more details, while at the same time preserving the highlights (which are already close to being blown). If you want that to happen (from an artistical POV) is another discussion.

With film photography we all tried hard to visualize our tones when we chose our exposure, surely that visualization process should apply to digital too. Exposure to the right is a great rule of thumb, but in terms of good RAW reproduction I don't believe it is the last word. There must be some sort of zone like system applicable to RAW, but I can only start guessing as to what that process might be.

You will always be limted by how much DR the sensor/cam can capture and to how much lattitude you have in PP.

jeremy_clay wrote:
Aside from the fact that problems at low-ISO without pushing the files are present (see the night shot noise on page 3 of the 'show us your shots the 5DII wrecked thread) The point many on the "Pro-5DII" side don't talk about is the fact that a regular 5D, the camera the 5DII was supposed to improve upon, is so much more flexible to be pushed in post without pattern noise. If they could do it before, why not now?


Because a sensor with larger pixels will almost always provide more latitude in the RAW image than a sensor with smaller pixels. This is one reason why there's still a market for cameras like the 10D, 1D Mark 1 and the original 1Ds (one of the sharpest and most detailed of all the Canon sensors to date). This is why the Nikon D700 provides more latitude than the D300 (same resolution; one camera is full frame with large pixels, the other is crop camera with smaller pixels).

Except for the above and the significantly higher resolution of the 5D II, whatever other reasons why the 5D II doesn't allow as much "push" as 5D images aren't known to me.

- Steve

pingflood wrote:
RobDickinson wrote:
Raw thesedays out of most SLRs is 11,12 or even 13 stops DR, prehaps not all usable but still. Film is surpassed, completly.

Yeah, many have this bizarre notion that print film is vastly superior to digital still. Often I hear that digital has "6-7 stops" range and color print film "14-15". Neither is even remotely correct in my experience.

Now, well processed black and white film (e.g. Tmax 100) I think still has the edge by a fair margin.

First, there is confusion between "lattitude" (how far away from an "ideal" exposure you can be without the difference becoming apparent in the print) and "dynamic range."

Second, when people speak of the dynamic range of print film, they are usually talking about what you can get when you pull all the stops on exposing for the shadows, developing for the highlights, carefully selecting paper grade, dodging, burning, et cetera.

With color negative film, their concept of dynamic range is more about latitude (how much the lab can correct poor exposure) than about dynamic range.

If you similarly pull all the stops with digital--proper exposure, post-processing, et cetera--you can certainly do at least as well as color negative film.

RDKirk wrote:
If you similarly pull all the stops with digital--proper exposure, post-processing, et cetera--you can certainly do at least as well as color negative film.

No question of it ... including the 5D II ....

- Steve

Except for the above and the significantly higher resolution of the 5D II, whatever other reasons why the 5D II doesn't allow as much "push" as 5D images aren't known to me.

Interestingly, we had an analogous situation with film. The higher resolving films had shorter dynamic range (and less latitude) than the lower resolving films.

I do think people are expecting a bit too much here.

Serious digital cameras have been around for what, less than a decade? Yet they are already matching if not beating medium format transparency film for sharpness, contrast, dynamic range and particularly high ISO ability.

Coming from an almost exclusively transparency film background (it was a treat for me to get a job I could use colour neg for so I didn't have to be so exposure critical) I have always used the same philosophy with digital - fill in the shadows if needed, use reflectors, keep the contrast down, try and expose to within 1/3 of a stop of correct exposure

The results have been just great, but I'm not expecting miracles out of the gear. I think some here are. I think David's shot is a great example of what dslr's can do.

David

Navyblue wrote:It's something that we use to wrap sandwiches.
hmmm ... Swine Flu Chop Sandwiches?

@Topic .. I too have read that film has much more DR compared to digital, and wondered why is it that we see better and better photos taken with digital cameras than scans from film. I guess the scanner is also to blame, but ...

RDKirk wrote:
Except for the above and the significantly higher resolution of the 5D II, whatever other reasons why the 5D II doesn't allow as much "push" as 5D images aren't known to me.
Interestingly, we had an analogous situation with film. The higher resolving films had shorter dynamic range (and less latitude) than the lower resolving films.

I never knew this. I always shot very low ISO print film and transparencies (usually ASA 64 or 25 for the latter!). This is useful information when trying to gauge digital.

- Steve

saaketham wrote:
I guess the scanner is also to blame, but ...


Which is digital too

I don't get people saying "Why could I push shadows more with the 5D when the 5DII is supposed to be better?"

5DII is a different camera. It has roughly double the MPs, smaller pixel density, better high ISO performance, video capability, etc.

I agree some people are expecting a perfect camera. Maybe we'll have that in 10-15 years, but for now the 5DII is a dream for me.

RDKirk wrote:

With color negative film, their concept of dynamic range is more about latitude (how much the lab can correct poor exposure) than about dynamic range.

If you similarly pull all the stops with digital--proper exposure, post-processing, et cetera--you can certainly do at least as well as color negative film.


Ok about the latitude, thanks for the correcton, not so sure though about getting more out of digital than film

dhphoto wrote:
The results have been just great, but I'm not expecting miracles out of the gear. I think some here are. I think David's shot is a great example of what dslr's can do.

I agree

Current DSLR's give great results. Although some do beter than others.

Take a look at this D90 sample. Perfectly exposed for the highlights. Meaning, there is still detail in the stained glass. Since I was in a tomb (US cemetary Normandy), I couldn't use flash.

In PP (LR 2.5) I adjusted the fill light by a whopping +60!!! This is the result: